Hall sensor VIOUT voltage when current sensed = 1.5A

[Flyback]

Hello can you confirm that when the current sensed in the ACS723LLCTR-10AU-T Hall sensor is 1.5A, the outputted voltage (VIOUT) is 600mV. (ie not 600mV +0.1xvcc = 1.1v)

ACS723 datasheet
http://www.allegromicro.com/en/Prod...p-Integrated-Conductor-Sensor-ICs/ACS723.aspx

 

[kubeek]

No, the output is I*0.4ohm+0.1*Vcc, which for I=1.5A and Vcc=5V is 1100mV.

 

[ronsimpson]

The 10AU version: Zero current = 1/10 supply.
0.4V/1A + (Vcc/10)
Range 0 to 10A.

The 5AB version: 0.4V/1A with and offset of 1/2 supply. Zero current = 1/2 Vcc. Range=_5 to +5A

 

[alec_t]

Bear in mind too that, as those specs show, the "Vcc x 0.5" is only a typical value, with no min or max given. So the output will need to be calibrated.

 

[Flyback]

sorry but for the unidirectional one of the top post, the "0.1 x vcc" that gets outputted for zero current is just due to the output opamp not being able to take its output down to 0v........the "0.1 x vcc" is not stated as being an offset, so for a sensed current of 1.5A, the output would be just 0.6V, surely?

In the bidirectional case, with a single supply, the output does need to have the permanent 2.5V offset so that you can distinguish between current flowing in either direction, so that is surely a different case.

The datasheet on page 8 states that the actual offset voltage is only 15mV. -Noting that it is not calling the "0,.1 x vcc" an 'offset'...because it does not come into play when the sensed current is above 1.25A, surely?

 

[kubeek]

No, if for any current between 0A and 1.25A the output stayed at 0.5V and then went from 0.5 to 0.6V between 1.25 and 1.5A the thing would be a very lousy and useless device. The offset is there precisely to get around the limitations in the opamp, not an effect of saturation and bad design.
Page 8 says offset is 0.1*Vcc, and accuracy on top of that is +/-15mV.
See page 18 for explanation of their terminology, however I am sure you have already read that

 

[Flyback]

Thanks but page 8 doesnt call "0.1 x vcc" an "offset"...it just says its the zero current output voltage....and it is surely simply the fact that the opamp cannot give an output at its low rail.
It doesnt make it a lousy device......for our application, it is perfectly fine that it gives 500mV for 0 to 1.25A, and then gives 400mV/A for 1.25A and greater.

 

[kubeek]

....and it is surely simply the fact that the opamp cannot give an output at its low rail.

I bet you that it is not. Get a sample and measure it if you can´t comprehend the datasheet.

Can you explain to me why would anyone want to produce such a crippled device? It makes no sense what so ever. No one will buy a device that starts working above 1.25A, and you are a one in a million case where such a behavior is useful.

And what about the bipolar devices? Do you think the 0.5*Vcc zero current output is also a manifestation of a screwed opamp, i.e any current from minus inifity to plus 6.25A is thrown away and anything above 6.25A starts showing on the output? Come on, you can´t be serious...

 

[alec_t]

I agree with Kubeek, post #2. As per the datasheet for the unidirectional IC the output voltage just ramps up linearly at 400mV/A from a starting point of 0.1 x Vcc.

 

[eTech]

VIOUT is about 1100mv. After removal of offset, its about 600mv (shifted output on graph).
I also get the same measurements on the bench.
See attached.

BTW- LTspice XVII has been released.

 

[Flyback]

Thanks, it would be nice to see what you get when you ramp the current from 0 to 1.5.
I looked for a spice model of it but couldnt find it.

 

[eTech]

Thanks, it would be nice to see what you get when you ramp the current from 0 to 1.5.
I looked for a spice model of it but couldnt find it.

Hi.

Attached is a simulated DC sweep from 1mA to 1500mA.
I'm also attaching a zip file containing the ACS723 series sensor models and a test circuit.
These are behavioral models I made for LTspice (Allegro didn't have them).
For the sim, I used the LTspice bundled universalopamp2 but for the bench I used an LM358.
I got very close results on the bench as I did on the sim.

BTW...I didn't model the BW pin, temperature, or magnetic coupling characteristics.

Have fun.

 

[Flyback]

thanks, oh right, i thought you were simulating the ACS723 module provided by Allegro...so the one youre simulating is one you made for yourself, to behave as you believe it should behave?

 

[eTech]

thanks, oh right, i thought you were simulating the ACS723 module provided by Allegro...so the one youre simulating is one you made for yourself, to behave as you believe it should behave?

Yes.

I tried to get an evaluation board from allegro but they don't have them yet..

 

[ronsimpson]

so the one youre simulating is one you made for yourself, to behave as you believe it should behave?

I have learned much by:
1) Reading the data sheet
2) Making the spice model
3) Turning on the real part and seeing how it works
4) Re-reading the data sheet.
5) Re-making the model
6) go step 3 (loop 10 times)

This is much like theology.
If you think things work (like this) it is easy to get the data sheet to agree with you.
A second person thinks things work differently and the data sheet agrees with them.
Likely neither is right.